Digitala Vetenskapliga Arkivet - Academic Archive On-line
Not a member yet
734120 research outputs found
Sort by
Machine learning-driven investigation of environmental effects on dynamic behavior of railway noise barriers based on long-term field test
No full textThe passage of trains by railway noise barriers induces vibrations that may affect their fatigue performance and reduce their service life. However, long-term field monitoring of noise barriers under complex environmental and operation conditions remains rare. This study develops an interpretable machine learning (ML) framework to investigate the aerodynamic pressure and dynamic behaviors of noise barriers based on a nine-month long-term field monitoring campaign, yielding 12810 train runs over 105 valid days. Input variables include train type, speed, temperature, wind speed and direction, relative humidity, and air pressure, while the target responses cover train-induced aerodynamic pressure, stress near the base of the steel post, and displacement at the post top. Eight ML models, including four traditional and four ensemble algorithms, were used and systematically compared to evaluate their predictive capabilities and robustness. Ensemble models, particularly Gradient Boosting Decision Tree (GBDT), Light Gradient Boosting Machine (LGBM), and Extreme Gradient Boosting (XGBoost), achieved the best predictive performance, with R2 values exceeding 0.935 for stress and displacement, and 0.895 for pressure. XGBoost, offering a strong balance of predictive accuracy and computational efficiency, was selected for SHapley Additive exPlanations (SHAP)-based interpretability analysis to uncover the physical relationships behind the data-driven predictions. Results reveal that aerodynamic pressure was the most challenging response to predict, given its higher sensitivity to turbulent airflow and environmental fluctuations, whereas stress and displacement exhibited more stable and predictable patterns. SHAP analysis identified train speed and type as the most influential factors across all responses. While environmental factors had comparatively lower influence, temperature and instantaneous wind direction consistently showed higher importance among them. Relative humidity has a moderate effect on aerodynamic pressure but a minor impact on dynamic behavior. Air pressure and wind speed exhibit limited influence on all outputs. These findings highlight the novelty and effectiveness of integrating long-term monitoring data, ML methods, and SHAP-based interpretability, offering new insights into the dynamic behavior of railway noise barriers.Validerad;2025;Nivå 2;2025-12-01 (u5);Full text license: CC BY 4.0</p
Developing and validating domain specific languages for cyberattack modeling and simulations
No full textThis thesis explores the potential of domain-specific languages (DSLs) to enhance the accuracy, efficiency, and expressiveness of cyberattack modeling and simulation. Motivated by the increasing sophistication of cyber threats, this work addresses the limitations of traditional modeling approaches by developing and validating two novel DSLs: one tailored for vehicular systems and another for the Information and Communications Technology (ICT) domain. These languages provide specialized vocabulary and syntax for describing attack patterns, system behaviors, and defense mechanisms concisely and straightforwardly. Through a series of experiments and case studies, this research demonstrates the effectiveness of these DSLs in capturing the complexities of real-world cyberattacks. These languages enable the automatic generation of attack graphs from system architecture models, streamlining threat identification and enhancing the alignment of security measures with established frameworks for cybersecurity professionals. This thesis contributes to the advancement of cyberattack modeling and simulation techniques, providing cybersecurity professionals with tools to express, analyze, and predict the behavior of cyberattacks.Denna avhandling undersöker potentialen hos domänspecifika språk (DSL) för att förbättra noggrannheten, effektiviteten och uttrycksfullheten i modellering och simulering av cyberattacker. Motiverad av den ökande sofistikeringen av cyberhot, adresserar detta arbete begränsningarna hos traditionella modelleringsmetoder genom att utveckla och validera två nya DSL:er: en skräddarsydd för fordonsystem och en annan för IKT-domänen. Dessa språk tillhandahåller specialiserad vokabulär och syntax för att beskriva attackmönster, systembeteenden och försvarsmekanismer på ett koncist och tydligt sätt. Genom en serie experiment och fallstudier visar denna forskning effektiviteten hos dessa DSL:er för att fånga komplexiteten i verkliga cyberattacker. Dessa språk möjliggör automatisk generering av attackgrafer från systemarkitekturmodeller, vilket effektiviserar hotidentifiering och förbättrar anpassningen av säkerhetsåtgärder till etablerade ramverk för cybersäkerhets-experter. Denna avhandling bidrar till utvecklingen av tekniker för modellering och simulering av cyberattacker, vilket ger cybersäkerhetsexperter verktyg för att uttrycka, analysera och förutsäga beteendet hos cyberattacker.QC 20251217</p
Biochemical osteomalacia reaffirmed by signs and symptoms and perinatal outcome : A prospective cohort study of women in Sweden
No full textBackground: Vitamin D deficiency-induced osteomalacia remains underexplored, despite the substantial migration to northern sun-deprived latitudes. In women, osteomalacia may impair smooth and striated muscle function and disrupt the birth canal. This study aimed to investigate the associations and effect estimates of biochemical osteomalacia on perinatal outcomes. Methods: A prospective cohort study was conducted to examine 71 Swedish and 52 Somali women during pregnancy and breastfeeding, addressing the heightened risk of severe vitamin D deficiency among Somali women. The baseline data comprised blood samples, questionnaires and clinical examination. Two years later, outcome variables were collected and comprised diagnostic codes for delivery methods. Women with miscarriage, stillbirth, or relocation from the region were excluded. Biochemical osteomalacia reaffirmed by signs and symptoms was diagnosed based on a non-invasive, non-radiation protocol. Associations between biochemical osteomalacia and delivery outcomes were analyzed using multinomial logistic regression, adjusted for a minimal set of confounders. Results: In the cohort 20 women, 19 Somali and one Swedish, were diagnosed with biochemical osteomalacia. Among women with biochemical osteomalacia, the adjusted odds ratio (aOR) for instrumental-assisted delivery was 4.92 (95 % CI 1.30-18.65) and the aOR for vacuum extractions was 16.16 (95 % CI 1.20-217.55). Conclusions: Biochemical osteomalacia was associated with a higher incidence of emergency instrumental delivery procedures, including an increased likelihood of vacuum-assisted delivery and emergency Caesarean sections. Primary healthcare staff play a vital role in screening for vitamin D deficiency during pregnancy and breastfeeding period, as well as initiating supplementation to mitigate the risk of adverse perinatal outcomes
Dynamic fractures in rock-like materials analyzed by high-speed photography and a dynamic phase-field finite element model
No full textA so-called hydrostatic-spectral-deviatoric decomposition is integrated into dynamic phase-field finite element models to study its effectiveness in the prediction of crack paths and crack tip velocities in brittle materials under dynamic mixed-mode compressive loading conditions. The study focuses on dynamic experiments, including drop tower tests conducted using gypsum plaster specimens with embedded flaws and holes. The hydrostatic-spectral-deviatoric decomposition method enables a split in tensile and compressive strain energy density contributions, crucial for capturing the complex fracture behavior in brittle materials under compressive loads that traditional models struggle to simulate effectively. Numerical simulations were conducted in parallel with experiments, utilizing a unified set of material parameters derived from prior work on the same material under quasi-static conditions. This approach ensured consistency between the dynamic and quasi-static cases. Numerical results demonstrated a good level of agreement with experiments, replicating key aspects of fracture evolution, including initiation and propagation of primary and secondary cracks as well as the temporal progression of crack tip velocities. This enhanced modeling approach offers a deeper understanding of dynamic fracture mechanisms in brittle materials and demonstrates its potential for advancing fracture simulations in applications involving dynamic, mixed-mode loading scenarios
Benchmarking foundation models and parameter-efficient fine-tuning for prognosis prediction in medical imaging
No full textBackground and Objectives: Despite the significant potential of Foundation Models (FMs) in medical imaging, their application to prognosis prediction remains challenging due to data scarcity, class imbalance, and task complexity, limiting their clinical adoption. This study introduces the first structured benchmark to assess the robustness and efficiency of transfer learning strategies for FMs compared with convolutional neural networks (CNNs) in predicting COVID-19 patient outcomes from chest X-rays. The goal is to systematically compare fine-tuning strategies, classical and parameter-efficient, under realistic clinical constraints related to data scarcity and class imbalance, offering empirical guidance for AI deployment in clinical workflows. Methods: Four publicly available COVID-19 chest X-ray datasets were used, covering mortality, severity, and ICU admission, with varying sample sizes and class imbalances. CNNs pretrained on ImageNet and FMs pretrained on general or biomedical datasets were adapted using full fine-tuning, linear probing, and parameter-efficient methods. Models were evaluated under full-data and few-shot regimes using Matthews Correlation Coefficient (MCC) and Precision–Recall AUC (PR-AUC) with cross-validation and class-weighted losses. Results: CNNs with full fine-tuning performed robustly on small, imbalanced datasets, while FMs with Parameter-Efficient Fine-Tuning (PEFT), particularly LoRA and BitFit, achieved competitive results on larger datasets. Severe class imbalance degraded PEFT performance, whereas balanced data mitigated this effect. In few-shot settings, FMs showed limited generalization, with linear probing yielding the most stable results. Conclusions: No single fine-tuning strategy proved universally optimal. CNNs remain dependable for low-resource scenarios, whereas FMs benefit from parameter-efficient methods when data are sufficient
Designing a regenerative pomegranate supply chain network under uncertainty: A Lagrangian relaxation approach
No full textRegeneration is a fundamental concept at the heart of the circular economy. This concept emphasizes increasing positive environmental impacts in addition to achieving zero waste. Inspired by this concept, in this research, a mixed-integer linear programming model is developed to design a regenerative supply chain (SC) network in the pomegranate industry under uncertainty. In the investigated network, all pomegranate waste is transported to composting centers. Then, the compost produced by the composting centers is made available to farmers, thus integrating the concept of regeneration with SC operations. A stochastic scenario-based approach is used to deal with the uncertainty of the demand parameter. In addition, a Lagrangian relaxation algorithm is applied in order to solve the problem in large sizes. The validity of the developed algorithm is examined by comparing its results with GAMS results on nine small-sized instances. The results show that the proposed algorithm provides solutions close to GAMS solutions while being faster. Finally, the applicability of the presented model is confirmed through its implementation in a pomegranate industry in Iran
Inclusive Playgrounds: Insights Into Play and Inclusion From the Perspectives of Users and Providers
No full textPlay for play's sake is an important part of a child's life. In this sense, play is also enshrined as a child's right and is understood from an occupational therapy and occupational science perspective as a central occupation in children's lives. Children report that outdoor environments, such as playgrounds, are some of their favourite places to play. However, studies also show that children’s experiences of play occupation in playgrounds, especially for children with disabilities, can be limited by barriers related to the physical, social, and political environment. To address these barriers, so called inclusive playgrounds have been developed and implemented with the aim of providing play and social experiences for all children to foster a sense of belonging and inclusion. Inclusive playgrounds could therefore be considered places created by playground providers for children where situational elements of the physical, social, and political environment converge with children's play occupation. The Transactional Model of Occupation (TMO) was chosen as the theoretical underpinning of the thesis with the aim of providing a framework for interpreting the perspectives of playground users and playground providers in relation to the intertwined nature of the situational elements from an occupational and child-centred perspective. Furthermore, the TMO was found to be useful in integrating other concepts related to inclusive playgrounds and their transactions with situational elements, such as play value, affordances, place-making, inclusion and Universal Design (UD). The overall aim of the thesis was to gain a deeper understanding on play and inclusion on inclusive playgrounds from the perspectives of playground users (children with and without disabilities and advocates of children with disabilities), and playground providers (including experts in Universal Design). The thesis was informed by four studies, whereby study I and study III looked at the children’s perspectives, study II at the perspectives of playground providers and advocates of children with disabilities, and study IV at the perspectives of experts in UD. Study I explored the experiences of children with (n=18) and children without (n=14) disabilities of playing on inclusive playgrounds through the use of interviews and observations. Data was analyzed using qualitative content analysis. Study III aimed to expand knowledge from a child-centred perspective of how environmental characteristics influence play value and inclusion for all children in outdoor playgrounds. The study was conducted as a meta-ethnography and included 17 studies. Study II explored the design and use of inclusive playgrounds with a particular focus on how design supports or hinders inclusion from the perspective of people involved in designing (n=14) or advocating for children with disabilities (n=12). Data were collected with focus groups and analyzed with thematic analysis. Study IV aimed to advance the understanding and use of UD in inclusive playground provision by identifying experts (n=6) strategies and experiences of applying UD in playgrounds. Data were collected with expert interviews using a go-along method of walk and talk interviews and analyzed with qualitative content analysis. The synthesis of the findings provided insights into children’s experience of participation in play occupation and play value on inclusive playgrounds; into how play value emerges from transactions of the situational elements; and into what UD adds to playground design to create a welcoming atmosphere and make playgrounds inclusive. Children’s experiences of play value were found to emerge from transactions of the play occupation and the physical and the social environmental elements, and sociocultural, and geopolitical elements, and leading to a sense of belonging. A sense of belonging was found to be associated with inclusion from the perspective of children and advocates of children with disabilities, as well as from the perspective of experts in UD. Thus, children’s perspectives on play value and participation in play occupation were found to contribute to an understanding of what makes a playground inclusive from a children's perspective. Furthermore, findings suggest that UD might be a useful approach for to design for inclusion in playgrounds, because it was found that for the UD experts, the social environmental elements and the sociocultural and geopolitical elements were at the beginning of the design process and guided the design of the physical environmental elements accordingly. This focus is also reflected in four strategies identified from the synthesis of the findings for designing playgrounds to promote a sense of belonging. To further explore play occupation and inclusion in playgrounds perspectives that look at communities rather than individuals, such as communal or collective occupations, may be useful to focus on the social aspects.
Calibrating ERAPave PP with field performance data
No full textThe conventional pavement design approach considers various parameters for the optimization of pavements for the prevailing traffic and environmental conditions. Demands related to technological developments, climate adaptation and infrastructure resilience are expected to influence the way pavements are designed and constructed. For this, better and improved flexible pavement design tools are required. ERAPave performance prediction (PP) which is a mechanistic-empirical (M-E) pavement design tool is currently under development with the goal of addressing the several challenges facing the pavement industry. This paper calibrates the permanent deformation prediction approach in ERAPave PP using pavement performance data from actual field pavements. As traffic volume is observed to have a significant influence on predicted results, separate calibration was performed for medium-to-high-volume and low-volume traffic categories. A global calibration factor is used for this purpose, significantly improving the accuracy of the prediction for both categories. Prediction accuracy can be improved further through the consideration of observed rut depth variability.
Unified Approach to Industrial Information
No full textIndustry 4.0, has significantly altered industrial processes by integrating advanced technologies such as Cyber-Physical Systems, Internet of Things (IoT), cloud computing and AI. Despite these advancements, the industry is facing challenges with achieving effective communication and interoperability. Through the wide variety of advances that has been made, an upstream of different heterogeneous data sources and protocols have emerged, even when studying a specific domain such as the industrial sector. It has led to fragmented interoperability and somewhat unreliable information exchange. This thesis presents research that explores the potential of ontologies and semantic modeling to address some of these challenges by providing explicit descriptions of concepts and their relationships, thereby enhancing a shared understanding and vocabulary, improving interoperability and stakeholder communication. Furthermore, efforts are made to enable system communication through OPC UA and the Arrowhead framework, to enable seamless interoperability. Despite the possible benefits of ontologies, challenges such as the need for experience and expertise in ontology development is required to created and maintain their reliability. Introducing the Industrial Data Ontology (IDO) as an industrial upper ontology, a newly adopted ISO standard, enables a higher level of knowledge abstraction. IDO describes industrial assets and processes throughout their lifecycle. The findings underscore the transformative potential of semantic models, ontologies, and seamless interoperability to enhance the quality of industrial information exchange and a more sustainable and reliable process. Future directions include exploring the integration of real-time data with semantic benefits, enhancing business transaction process, and implement semantic explicitness to a Service-Oriented Architecture (SOA) such as the Arrowhead framework